By Lee Flanagan
Chris Johnson has a specific kind of candidate in mind when he talks about AI application fraud, and it is not the résumé with a slightly inflated job title. It is someone “running two or more full-time W-2 jobs” who fabricates a resume, invents an identity, and onboards and gets paid under it. In Johnson’s words, that person “might actually be a fraudster or a hacker trying to get into your system.” Johnson, an SVP and general manager of employer services at Equifax, told HR Brew that story because it is the one that survives the sifting. The outlet’s framing asks whether AI has become both the solution and the problem in hiring. The sharper question is what happens to your hiring process once you accept that it has.
The Fraud Equifax Is Selling You a Fix For
Equifax’s own report, cited by HR Brew, puts numbers on the anxiety: 78% of people pros say AI increases hiring and onboarding efficiency.
Two more numbers cut the other way: 36% say AI-generated candidate information has reduced their confidence in hiring decisions, and 73% report seeing fabricated or misleading candidate information outright.
Those figures are worth taking seriously, and they are also worth reading for what they are: research from a company that sells identity verification and background-check services, describing a problem those services are built to solve. That does not make the data wrong. It does mean the report’s own prescription, more verification tooling after the fact, is the answer Equifax was always going to give.
Where the Checks Actually Sit
Bart Lautenbach, Equifax’s SVP and general manager of talent solutions, names three points in the hiring process where fraud should be caught: the interview, the background check, and onboarding. That is a more honest map than the report’s headline stats suggest, because it names the interview as a checkpoint the background check cannot reach. Lautenbach also flagged something the background check cannot touch: job seekers who have someone else interview on their behalf, a risk he tied directly to remote roles, since hiring managers and recruiters might never meet candidates face to face.
Once identity substitution happens inside the interview itself, a check run afterward authenticates paperwork, not the person who actually did the work.
Why Going Back to a Room Does Not Scale
The Wall Street Journal reported in late 2025 that McKinsey and Google have reverted to in-person interviews in response to this exact concern. It is the bluntest fix available: put the candidate in a chair where you can see them. Even Lautenbach does not treat it as universal, telling HR Brew “that doesn’t have to be the solution for every company.” He is right, and the reason matters.
In-person interviewing solves identity confirmation only for roles where geography, cost, and candidate pool allow it. Most hiring does not sit in that category anymore. Global sourcing and remote-first hiring are the norm now. In our view, undoing that to solve a verification problem trades a hiring-quality gain for a fraud-prevention gain that most organizations cannot afford at scale.
The Interview Was Built to Assess Fit, Not Verify Identity
The interview is where that swap happens. A background check cannot catch a stand-in who is never on file, because the check only runs after the interview is over. That leaves the interview doing double duty. It has to assess whether someone can do the job, and it has to confirm that the person answering the questions is who they claim to be, with evidence that holds up under follow-up.
Look at your own scorecard. It routinely tests for competency and cultural fit, but rarely includes a single question built to pressure-test whether a candidate’s stated experience is real rather than plausible.
A background check confirms a document. It does not confirm that the candidate sitting across from you today is the one who will show up on day one.
Lautenbach’s phrase for the right posture is “trust, but verify,” paired with his point that strong data combined with a strong process “allows you to expand as far and wide as you necessarily need to be.” That is a fair description of what identity verification technology can do before and after the interview. It is not a description of what happens inside it. Cross-checking IP addresses and credentials catches the fabricated resume. It does not catch the interviewer who never asked a candidate to walk through the specific decision they claim to have made, in enough detail that a fabricated answer would fall apart.
Who Owns the Gap
Equifax’s report is right that AI has changed the shape of the risk. It is less clear-eyed about where the responsibility for closing it actually sits. You cannot outsource verification entirely to a vendor whose checks only start once the interview has already happened, when nearly three-fourths of HR pros are already telling Equifax that fabrication is getting through. The interview is the last point in the process where a human is live in front of you, and it is the point least equipped, right now, to do the verifying. That is the imbalance TA has not closed: more tooling on the back end, and no comparable investment in training the people running the front end to ask questions a fabricated identity cannot survive.
Original reporting: HR Brew.
Frequently asked questions
Is Equifax’s fraud data credible given the company sells verification services?
Equifax sells identity verification and background-check services, so a report describing rising fabrication and falling confidence in hiring decisions doubles as a case for its own product line. The underlying numbers, including 78% citing efficiency gains and 73% reporting fabricated information, still hold up on their own terms.
How does the ‘someone else interviews for me’ scheme actually work?
Bart Lautenbach of Equifax says some job seekers arrange for another person to sit in on their interview, a risk tied to remote hiring where the recruiter and candidate may never meet face to face. Because the substitution happens live, it leaves no paper trail for a background check to catch.
Why doesn’t a background check catch a candidate who used a proxy for the interview?
A background check verifies that submitted documents and credentials are authentic. It says nothing about whether the person who sat in the interview matches the name on those documents, and by the time the check runs, that swap already happened live with no paper trail behind it.
Why isn’t going back to in-person interviews the standard fix?
Lautenbach himself said in-person interviewing does not have to be the solution for every company. It mainly works where geography, cost, and candidate pool allow face-to-face hiring, and most roles now run on remote and global sourcing.
Why does the interview carry more weight than the other two fraud checks Lautenbach names?
Lautenbach’s framework lists the interview, the background check, and onboarding as the three points to check for fraud, but only the interview happens live, in real time, with an actual human across the table. The other two run after the fact, on documents and data, which is exactly where identity substitution during the interview slips through.